: The investigators will design and field the third wave of a public-use longitudinal household (HH) and community-facility (CF) survey: the Indonesian Family Life Survey (IFLS3). The survey will follow some 7,500 households and roughly 321 communities in 13 provinces that were initially selected for the first wave, in 1993 (the IFLS1), and re-surveyed in late 1997 (the IFLS2). The proposed data will greatly facilitate the study of demographic and socioeconomic change pertaining to the elderly and the processes of aging over a period of time that includes both long-run economic growth from 1966 to 1997 and a sharp economic downturn since that has been repeated in numerous countries. Adding the IFLS3 will allow for the examination of the impact of the crisis on well being of the elderly: looking, for example, at such variables as living in poverty, intergenerational transfers, labor force transitions, health outcomes, and health care utilization. However, the IFLS3 is much more than a survey that will focus on crisis. It adds a third wave to one of the best of a very few socioeconomic panel data sets, with extensive demographic, economic, and health measures, in a developing country. When the IFLS3 is completed, prospective longitudinal data will be available on such factors as household consumption and assets, individual health outcomes, health care utilization, living arrangements and other intergenerational transfers for a period spanning almost a decade. In addition, the surveys are well known for collecting retrospective data on such outcomes as births, marriage, school investments, migration, and employment. The implementation of the survey will continue and improve upon on a number of field-related methodological innovations implemented during the previous surveys that dramatically raised the quality of the surveys, resulting, for instance, in very high successful tracking rates, 94 percent between the first two waves and in the ability to check data discrepancies in the held days after they have been collected.